Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Mobile Money Report


Published on

“Beyond payments – Next generation Mobile Banking for the Masses”

Published in: Business, Economy & Finance
  • Be the first to comment

  • Be the first to like this

Mobile Money Report

  1. 1. Table Foreword Our Findings in a Nutshell Introduc on Mobile Micro nance Pilots in West Africa and South East Asia of of 4 5 6 8 Key Challenges in Mobile Micro nance Deployments 9 The Solu ons: Strategic and Opera onal Learnings from Pilots 13 Concluding Remarks 29Contents t Beyond Payments „ Next Genera on Mobile Banking for the Masses 3
  2. 2. Foreword2 009 and 2010 were exci ng years for mobile money. As the number of mo- bile money subscribers worldwide neared a 50 Mn1, the market saw some prom- ising launches and MNO-bank partnerships2 with a poten al to further innova on in the eld of mobile nancial services. However, while mobile money is now on the agenda of several mobile network operators (MNOs) and nancial ins tu ons, most are s ll living an M-PESA3 dream, seeking to replicate the success of Safaricom’s mobile payments ser- vice. PlaNet Finance and Oliver Wyman looked beyond payments to explore the power of mo- Greg Rung Arnaud Ventura bile technology in providing low income con- Partner Co-founder, Vice president & CEO sumers with access to a wide range of nan- Oliver Wyman PlaNet Finance Group cial products through next genera on mobile micro nance services. With wide coverage of bo om of the pyramid While the success of M-PESA has been ap- clude the role of regula ons in shaping emerg- (BoP) consumers, 2nd genera on mobile mi- plauded, it is s ll unprecedented making some ing models of mobile money, and the role of cro nance products – such as micro-savings, investors and providers cau ously enthusias c technology (with regard to pla orms, devices credit and insurance – can transform banking, about mobile money. Indeed, such an under- and security) in making mobile money a ord- especially in countries with less than 10% re- taking comes with formidable complexi es able, convenient and e cient. We will exam- tail banking penetra on. Extending the mobile and the formula for success is not straigh or- ine these later, when we move from the pilot money o ering beyond payments can signi - ward. This is the core topic we address in this stage towards commercial launch. cantly increase its value for both consumers report. Drawing on our experiences in pilots and providers. conducted in Africa and in Asia, we explain the We would like to thank all the experts who challenges in deploying mobile micro nance have contributed to this report through the and o er strategic and opera onal solu ons. direc on and views they have provided, es- In doing so, we have explored two dis nct in- pecially Daniel Radcli e and Ignacio Mas from nova ve models through our pilots. the Bill & Melinda Gates Founda on and Mark Flaming, an interna onal expert in micro - The distribu on of micro nance through nance and mobile banking. We hope that this mobile money paper will provide insights into issues that prac oners across markets encounter when A virtual micro nance bank, a pure deploying innova ve mobile money solu ons. mobile play without any branch based banking opera ons It should be taken into account that our learn- ings are limited by the scale and scope, both in size and characteris cs, of the pilots de- scribed in the report. Some important aspects fell outside the scope of this study. These in- Greg Rung Arnaud Ventura1 Joint study conducted in 147 emerging markets by GSMA and CGAP.2 Such as Safaricom/Equity Bank in Kenya, Tameer Micro nance Bank/Telenor in Pakistan, SBI/Airtel in India.3 M-PESA is the mobile money transfer service launched by Safaricom in Kenya. It is among pioneer products of the mobile money world.4 Beyond Payments „ Next Genera on Mobile Banking for the Masses
  3. 3. Our Findings in a NutshellI B n this ar cle, we introduce “mobile micro- elow are our key learnings for se ng up nance” – a way of providing micro nance mobile micro nance, based on observa- 5. Marke ng communica ons for a mobile products through the distribu on network ons from our eld tests: micro nance enterprise need to address twoof a mobile operator – and “virtual micro- key issues. nance bank” model, where the provider 1. The strategy alignment phase between 5.1. Given low nancial literacy in targetserves its customers only through the mobile partners can o en be long and can signi - segments, it should s mulate trust and sen-channel, without deploying brick and mortar cantly delay launch melines. Partners should si ze consumers towards nancial products. Abranches4. have similar appe tes for the me and invest- well known brand can a ract customers to try ments required and should agree upfront on a the service but it does not guarantee sustainedThe bene ts of these models, observed in our high-level strategy, while kicking o execu on usage. Brand reposi oning and further trustpilots, include a more than twofold increase in in parallel. building measures might be needed at bothaccess to banking, 20-50% lower opera onalcosts for the MFI and revenue or market share 2. It is o en di cult for partners to agree product launch (through above the line cam- paigns) and at customer acquisi on (throughbene ts for the MNO. The main conclusions of on a sustainable pro t model. Nego a ons can be eased by on-going discussions support- street marke ng ini a ves).our study are: ed by detailed customer valua on exercises„ Vanilla micro-savings have very low usage and exclusivity to protect early investments. 5.2. Op mizing marke ng spend can be a rates. Introducing commitment driven bundled challenge in serving low-income consumers. savings5 has been effective in stimulating usage The model should be both ac va ons and us- Partners can maximize the marke ng dollar in our pilots, boosting activity rates by 2 to age based and should allow for one ac vity or impact by aligning marke ng campaigns with 3 times. If the bundled product eliminates a product to subsidize the other. segments’ nancial cycles and leveraging vi- “pain point” for the consumer, it can catalyze the movement of deposits from “under the 3. Product design for such an enterprise ral marke ng techniques to boost ac va ons (through events, agents and peer-to-peer in- mattress” to a bank. poses two key challenges.„ There is typically a high seasonality of savings terac ons). among different sub-segments and partners should consider this in their product design and 3.1. Poor understanding of the nancial 6. Building, managing and incen vizing an in deciding when to market products. needs of target segments a ects product se- MNO’s distribu on network in tandem with„ A meticulous performance management system quencing and design. an MFI’s agent network is a complex under- can have a significant effect on agent behavior. When implemented correctly, we have seen such „ Choice of launch product should be based taking. We have tested “Mobile Coaches” as a system increase activity rates by up to 60%. on the market’s unmet needs and readiness a way of integra ng MFI-MNO’s networks for mobile money. The presence of 1st and boos ng ac va ons using feet-on-street generation products can be a precursor to agents. Partners should also: launching 2nd generation services. „ Adopt a systematic selection norm to sign „ Selectively targeting sub-segments and up cash-in/out agents that most likely do adapting offer design to their income cycles well on financial products. can reap benefits. „ Implement a value-based incentive scheme and a performance management system 3.2. Most products face very low usage for agents while also subsidizing the agent post-acquisi on due to earnings vola lity network to ensure an attractive business for and a lack of nancial discipline. For savings, them. a combina on of commitment to save (from the consumer) and a bundled incen ve (that 7. Operators must nd ways to manage drives savings rates) can be an e ec ve way to risks that mul ply as a small scale MFI ex- s mulate deposits. pands services to a much larger network through remote channels. MFIs should imple- 4. Poor customer experience is a key rea- ment an automated risk management system son for dropouts: partners should create a with four key levers: data-driven credit6 scor- simple and consistent customer journey to re- ing, gradua ng credit, predic ve loan man- enforce trust, ensure access, facilitate usage agement systems and use of borrowers’ vital and maximize customer life me value. business par cipants in distribu on7.4 There are branches in the form of cash-in/out agents, but they do not have to form part of the formal banking system.5 Savings coupled with other products such as credit/insurance/high interest.6 Gradua ng Credit: When the current loan amount for which the client quali es increases based on a metric that can enable assessment of the clients credit worthiness e.g. repayment performance of previous loan.7 Suppliers, creditors, clients etc. Beyond Payments „ Next Genera on Mobile Banking for the Masses 5
  4. 4. Introduction ‘‘ 2nd generation mMoney products can benefit billions of people by giving them access to basic microfinance/banking ser- vices while creating a profitable business model for operatorsT he number of unbanked or under- banked mobile subscribers around the world is projected to reach ~2 billion by 20128. Today, only around 50 million9 sub- surance, especially in countries with less than 10% retail banking penetra on. A progressive take-up of more such complex mobile banking products will allow both telcos and nancial ‘‘ exis ng levels11. In one of our pilots, 80-90% of the acquired mobile micro nance subscribers had never held a bank account previously. A widespread “Mobile Operator like” distribu- scribers use mobile money services. This busi- ins tu ons to bene t, as they reap value from on model makes the service far more acces- ness has seen huge growth in recent years, complementary skills and deliver more value sible and convenient over the exis ng branch- both from mobile network operators and to customers, as illustrated in Figure 1. This based banking op ons, reducing travel and nancial ins tu ons, with 93 live and nearly report focuses on “mobile micro nance” as a opportunity costs for consumers. In our pilot 100 addi onal planned deployments10. Most way to reach a large mass of unbanked sub- in South East Asia, these savings were approxi- deployments so far have been focused on scribers through not only payments, but also mately 80%, as shown in Figure 2. These ben- 1st genera on mobile money products such more advanced banking products. e ts come together with the well-established as remi ances, air me top-up, bill payments advantages of tradi onal micro nance which T and loan repayment. The transforma onal he upsides? Mobile micro nance is plays an important role in stabilizing the cash- impact of mobile money is expected to come set to bring signi cant bene ts for all ow of low income segments. from 2nd genera on nancial services such stakeholders. For clients, it can expand as micro-savings, micro-credit and micro-in- access to banking by more than twice over the Figure 1 * MNO/bank value creation potential along customer migration path High Payments Savings Lending Customer Revolving path Credits Lending Savings Value added for mobile Cards Customer value added operator Retail purchase Service sfer Bills tran Salary Value added dit for bank cre Accrual bile Mo / Z P2P Money Transfer Cash Withdrawal P2P Cash Money Deposit p up Transfer Mobile To Low High 8 Joint study conducted in 147 emerging markets by GSMA and CGAP. 9 GSMA Mobile Money Deployment Tracker 2010; 45 Mn Mobile subscribers across di erent deployments. 10 GSMA Mobile Money Deployment Tracker 2010. 11 Gap between banking penetra on @10-30% and mobile penetra on @60-80% in the pilot geographies.6 Beyond Payments „ Next Genera on Mobile Banking for the Masses
  5. 5. F or Banks and MFIs (Micro nance Ins - scriber base by 60% by cross-selling through tra ve and personnel costs owing to the use tu ons) the delivery of nancial prod- the mobile channel. In our South East Asian of a purely remote delivery channel through ucts through the mobile channel can case, the mobile MFI has a poten al to reach largely automated processes.allow MFIs to move further down the income up to 10 mes the subscriber base of a typical Even for a tradi onal MFI deploying mobile aspyramid, increasing their client base many- MFI in the country. an alterna ve delivery channel, cost savingsfold. In the case of Kenya, Equity Bank aims The mobile channel comes at a lower cost than can come from an op mized distribu on mod-to double its number of bank accounts over branch based banking. For the “virtual MFI” in el and automated or remote client handling.a period of 15 months12 by reaching remote South East Asia, the opera onal costs are 50% In our pilot in West Africa, the MFI reduced di-customers partly with the help of their mobile lower than those of a tradi onal MFI, as illus- rect transac on costs up to 37% and process-savings product. In our case studies, the MFI trated in Figure 3. These savings come from a ing costs by around 25% by implemen ng ain West Africa can poten ally increase its sub- very lean branch model with lower adminis- mobile-based loan repayment system14.Figure 2 * 13Cost for a savings client of the virtual MFI (in the pilot) vs. traditional MFI(Basis points on a base of a 100; for a 50 USD deposit over 26 weeks)Source: South East Asia Pilot Cost to client Cost di erence 120 100 80 53 60 100 40 30 20 21 0 -4 Travel cost for Saving in travel cost Saving in opportu- Travel cost for mobile MFI D&/ to client nity cost to client fee (withdrawals)12 Reuters; Equity bank press release, Oct 4, 2010.13 Comparison drawn between the branch PnL of the virtual MFI and a comparable tradi onal MFI o ering same product lines(individual loans and savings as against group loans/savings) to low income segments; costs adjusted to di erences in labor and resources costs in markets of comparison.14 For some tradi onal banks/MFIs, stated costs savings might be partly o set by the expenses incurred in modifying and adop ng the bank’s legacy systemsto interact with MNO’s advanced mobile money pla orms. Furthermore, running branch based and mobile channels in parallel can dilute these cost savings on the overall cost base. Beyond Payments „ Next Genera on Mobile Banking for the Masses 7
  6. 6. Mobile Microfinance Pilots in West Africa and South East AsiaF With funding from the Bill and Melin- or Mobile Network Operators (MNOs) phone, regardless of their loca on. Second, there are also clear bene ts. Some can da Gates Foundation, PlaNet Finance it o ers a product that allows customers to be quan ed, such as increased market and Oliver Wyman worked together in repay their loans remotely, using a mobile- share, reduced customer churn15, enhanced early 2010 to explore the field of mo- banking pla orm. Third, it develops new retail revenue per subscriber and increased revenue bile microfinance via pilots in West points for mobile money opera ons, thus giv- from commissions. In our pilot in South East Africa and South-East Asia. ing easier access to products for its customers, Asia, more than 90% of subscribers acquired while genera ng added revenue and new cus- O ur pilot in West Africa is a partnership by the mobile MFI are new clients of the tomers for the MFI. between one of the leading MNOs in MNO’s mobile money service and hence gen- Our pilot in South East Asia is a partnership the con nent and a local tradi onal erate added revenue from commissions. There between a leading domes c MNO and a start- Micro nance Ins tu on. The project has are also broader bene ts that cannot be fully up branchless bank. The key objec ves are to: three main goals: quan ed at this stage but include strengthen- „ Increase access to core banking products „ Fight poverty by enabling more people in ing the MNO’s brand posi oning and reinforc- for low income consumers living in remote low income segments to build up savings. locations. ing the MNO’s appeal to clients and investors. „ Promote Micro-Entrepreneurs by making „ Increase banking convenience by enabling Of course, these bene ts do not come through credit more affordable. people to save money by going to outlets easily. Launching mobile micro nance is a ma- „ Make life easier for the unbanked by bringing near their homes. jor transforma onal e ort, both for tradi onal microfinance products to mobile phones. „ Develop a low cost and competitive nancial ins tu ons and for MNOs. We ad- The MFI aims to achieve these goals in three operating model with financial gains for dress the key strategic and opera onal com- ways. First, it allows customers to make de- the Bank and its distribution partners (the plexi es in subsequent sec ons of this report. posits on a saving account using their mobile MNOs and 3rd party distributors). Figure 3 * Cost to serve savings client for the virtual MFI (in the pilot) vs. a traditional MFI (Basis points on a base of 100; for a 50 USD deposit over 26 weeks) Source: South East Asia Pilot Cost to bank Cost di erence 120 100 80 43 60 7 100 40 50 20 0 Travel cost for ^avin in ed ^avin in ac isi on Cost for obile D&/ tradi onal D&/ branch cost and servicing cost 15 The rate at which customers leave for a compe tor.8 Beyond Payments „ Next Genera on Mobile Banking for the Masses
  7. 7. Key Challenges in Mobile Microfinance DeploymentsP artners in this model aim to set up a “vir- design. Our main learnings are outlined below. ‘‘ tual Micro nance Bank”. This Bank will Complexities in deploying a Many factors a ect the development of mobile nancial services. While a model may succeed use the MNO’s exis ng agent network to successful mobile microfinancedistribute nancial products to unbanked cli- in one region, it may struggle in another due to enterprise are many; operatorsents. It does not intend to own a conven onalbanking arm or brick and mortar customer ser-vices branches. The key principle at which such need a holistic strategy that carefully addresses all the deli- cate elements of the business ‘‘ di ering nancial needs or behavior of target subscribers, the maturity of the banking and telecom markets, strategies adopted by provid- ers or regula ons. For example, M-Pesa wasa virtual micro nance bank operates is that not ini ally as successful in Tanzania as in Ke-several of its core func ons are outsourced model nya because of di erences in the mobile bank-and the key task of the bank’s human resourc- ing landscape in the two is to manage partners be it for distribu on, We engaged with the partnering banks and Given the capabili es required to o er mobileclient servicing or data management. The MNOs in four countries over a period of 8 nancial services and the regulatory regime inbanks’ o ces do not serve as cash-in/cash-out months to support various ac vi es including most markets, a mobile micro nance enter-branches but purely as partner service centers strategy re nement, business planning, mar- prise can be delivered only through strategicand most of the ac vi es are automated and ket research, product design and tes ng, dis- partnerships. Some of the partnership op onscentralized. tribu on strategy design and opera ng model are illustrated in Figure 4 below.Figure 4 *Mobile Microfinance: Partner models tested in pilotsSource: Pilots and Emerging Mobile Money Models “Virtual Bank” “Hybrid MFI” Mobile Network Mobile Network Mobile Network Operator(s) Operator(s) Operator(s)16 CGAP: Mobile Banking in Tanzania, July 2009: This di erence is a ributed to three main factors: 1. Tanzania is more geographically spread vs. Kenya which is densely populated, making it di cult to op mize agent outreach. 2. Vodacom in Tanzania faced s compe on from exis ng mobile payments providers when it was launched. On the other hand, Safaricom was the dominant player with more than 2/3rds of market share in Kenya. at the launch of its mobile payments service. It was also one of the rst providers of mobile payments. 3. Vodacom Tanzania had a lower control over its agent network than Safaricom in Kenya and also had signi cant di erences in its posi oning compared to Safaricom in Kenya. Beyond Payments „ Next Genera on Mobile Banking for the Masses 9
  8. 8. W e have subse- quently split the business model of a mobile micro- nance enterprise into six key func ons which need to come together in harmony to ensure a successful mix. These func ons are illus- trated in Figure 5. Each of these plays an important role in the delivery of servic- es; however some are more complex than others. We will focus in the coming sec ons on the seven broad areas that posed the highest implementa on challenge or emerged as key strategic levers in our pilots. Some of the elements fell outside the scope for tes ng in the pilots, owing to a rela vely limited scale of opera ons. Figure 5 * Business model elements of a mobile MFI Source: Assessment of Pilots in South East Asia and West Africa BAA#1 Partnership Set-up The e ort and me required to align partners fully on strategic areas can signi cantly delay project launch, pu ng success at risk. The me taken in aligning the strategy and agreeing on the way forward was signi cant in both our pilots. Partners took almost a year to reach advanced stages on strategy design. Some of the most common issues encoun- tered in strategy design were choice of prod- ucts, pricing, target segments and branding decisions. In terms of strategy execu on, part- ners made a signi cant number of itera ons in 17 18 nalizing the scope of work and the data man- agement approach. The most common reason for this issue is a lack of understanding of the customer and of the partner’s assets, and how they can best be leveraged. For instance, in the case of the West Africa pilot, the MFI already had ten exis ng and planned products, along with its own money transfer o er, which need- ed to be ra onalized alongside the MNO’s mo- bile money products upon the launch of the mobile micro nance partnership. #2 It is o en di cult to agree on a pro t model that is sustain- able for the MFI while being su ciently a rac ve for the MNO. The pro t model directly a ects the pricing for the consumer and is hence a very cri cal ele- ment of the partnership. O en mul ple itera- ons are required before xing both the basis for pricing as well as the levels. This happens mainly due to a lack of understanding of cus- tomer economics before the pilot, especially if the service is new and the adop on rates can- not be accurately determined. Given the studies conducted are in a pilot stage, operator-bank rela onship structuring (par cularly forms of partnership and related legal, commercial, opera onal terms) and governance (par cularly management con- trol and decision making) have been excluded from the scope of the paper. Interested readers can refer to the recommended publica on: ”Mapping and E ec vely Structuring Operator-Bank Rela onships to O er Mobile Money for the Unbanked” by Neil Davidson, GSMA Mobile Money for the Unbanked, which elaborates this topic in detail. While Regulatory and Licensing can pose important challenges in many markets, given the broad range of issues involved in this, we have le it outside the scope of the current paper. Interested readers can refer to the recommended publica on “Regula ng New Banking Models that can Bring Financial Services to All” by Claire Alexandre, Ignacio Mas and Dan Radcli e from the Bill & Melinda Gates Founda on, which elaborates this topic in detail.10 Beyond Payments „ Next Genera on Mobile Banking for the Masses
  9. 9. BB#3„ Marketing Product design for such an enterprise poses two key challenges. 3.1 The financial behavior and pain points of low income segments are not well understood, making it tricky to choose and design suitable products.The foremost ques on is that of product se-quencing, in other words selec ng whichof the rst and second genera on productsshould be o ered in a given market. For in-stance in our West Africa pilot, we found thatthe uptake for the mobile money solu on wasrela vely low, but the MFI was su ciently so-phis cated to launch complex banking prod-ucts. Hence, we rst introduced the MFI tosimple mobile payment products to give boththe MFI and its customers me to learn.A subsequent challenge is to develop an o erthat is well suited to the needs of low incomesubscribers. Most of these subscribers havebeen largely cash dependent and their nan-cial behavior is very unsophis cated. Inad-equate upfront investments in understandingthese subscribers can result in the design ofproducts that don’t t into their nancial life. #4 Poor customer experience can result in drop outs a er ini al usage of products. We have seen mul ple factors that can lead to a loss of the client’s trust a er ac va on. In our pilots such factors included delay in SMS con rma on a er deposit, inconsistent expe- rience with di erent cash-in/cash-out agents, a di erence in balance tally between client’s passbook (self recorded) and the bank’s state- ments, due to manual recording mistakes and other factors. In the case of the West Africa pi- lot we saw transac on failure rates as high as 30%. Most of these subscribers were not using the service subsequently. #5 „ Marke ng communica ons. 5.1 Given low financial literacy in “Cash Societies”, a key role of the marketing communications in launching an innovative financial product is to build trust and to raise consumer awareness. The target segments of mobile micro nance services have typically had very limited expo- money transfer service, due to the wide recog- ni on of the MNO’s mobile payments brand. Moreover, we observed that even in periods when consumers stated that they saved more than 30 USD per month, they were only put- ng small por ons of that (5-10 USD) into the mobile banking account, indica ng limited trust in the service. These examples show that nancial awareness is a key role for market- ing func ons and that the posi oning of the brand needs to take this into account. „ 5.2 At early stages, on a small customer base, optimizing marketing spend can be a challenge. Most o en, MNO’s mobile money business models mainly rely on person-to-person re- mi ances, while other mobile micro nance services are not su ciently substan al to jus fy large scale communica on budgets. Hence, MFIs need to nd ways of acquiring subscribers at a low cost and devise e ec ve marke ng tools to maximize ac va ons and usage. In one of our pilots, a cost assessment exercise revealed that if marke ng expendi- tures of the virtual MFI were as high as a con- ven onal mid-large scale MFI19, the subscriber acquisi on cost would make most customers loss-making, given the exis ng transac onFor instance, we no ced in our pilot in Asia, sizes. Se ng a higher minimum transac on sure to nancial products before. The bene tsthat the rst savings product were unable to size, on the other hand, would signi cantly and advantages of formal banking are notmobilize client’s deposits despite a heavy mar- reduce ac vity levels. Given limited marke ng therefore apparent to them. In one of our pi-ke ng push, simply because the product did budgets dedicated to the launch, this situa on lots we noted that 60-70% of the mobile sav-not address the client’s nancial pain points. prompted the MFI to devise ways to op mize ings subscribers ini ally thought of the mobile their marke ng spend for maximum impact.„ 3.2 Most products face very low usage micro nance product as an air me top-up or a rates post-acquisition.Customer acquisi on can be easy for a servicethat is provided for free. For example, freeac va on of savings accounts in our pilotshelped the agents acquire up to 60-70% oftheir leads. However, sustained usage of theservice has been the biggest challenge in ourcase studies. In case of the West Africa pilot,less than 5% of the customers were ac veini ally. Similarly in the South East Asia pilot,around 70% of the acquired subscribers onthe rst savings product never used their de-posit accounts for any transac on. This largelystems from the fact that low income segmentso en have highly variable incomes, very basicknowledge of nancial products and most arenot disciplined savers. We no ced that almost20-30% of subscribers in the pilot were dor-mant due to a lack of understanding of howexactly to use the products. To overcome thisbarrier, tradi onal MFIs have maintained ahandholding approach, staying very close tothe customer, which might not be scalable orcost e cient for a sizeable customer base.19 Comparison drawn to a comparable MFI owned and run by the same conglomerate with ~40,000 subscribers. Beyond Payments „ Next Genera on Mobile Banking for the Masses 11
  10. 10. CC#6 undertaking. Sales & Distribution Building, managing and in- cen vizing an MNO’s distri- bu on network in tandem with an MFI’s agent network is a complex The agent networks of an MNO and an MFI are signi cantly di erent in structure, and the skills of the agents themselves are very di erent. This can become a key issue when deploying a joint network of agents for de- livery of mobile nancial services. Typically, the MNO’s agent network has a number of layers (not directly controlled by the MNO) with a large-scale presence and a high client- to-agent ra o. Agents have narrow agent skills sets (owing to the commodity nature of core mobile products). In certain MNO networks, the operator only exerts limited control over the network, mostly run by 3rd party distribu- tors20. The MFI agent network on the other hand is limited in scale; loan agents are highly skillful sellers as well as nancial tutors; they are used to close proximity to the customer, and most of the network is controlled directly by the MFI. This makes it di cult to translate MFIs also need to nd ways to scale up their agent networks at an op mal pace especially when the MNO’s mobile money networks do not provide the required capillarity. For instance, in one of our pilots, many of the MNO’s exis ng cash-in/cash-out agents (for remi ances) are in urban/semi-urban areas. Subsequently, the MFI expects to build up its cash-in/cash-out network for rural areas from scratch. the en re MNO’s agent network into a mobile money distribu on network. In addi on, the commission levels on telecom products can typically be much higher in both amount and volume, compared to nancial products which ini ally have low uptakes. Such di erences make co-managing the two systems a major challenge. We faced this chal- lenge in pilots where the MFI encountered sig- ni cant resistance from air me sales agents in taking on oat management responsibili es, owing to a lack of me and low commission incomes from mobile savings products. D Risk Management D#7 Operators need to nd ways nance Bank in our pilot study aims to provide however to be compensated with dynamic to manage risks that al- loan products, disbursed on the subscriber’s and automated credit risk management, by low a small scale MFI to ex- mobile wallet in a completely automated and way of reminders (via automated SMS), data centrally controlled way. Tradi onal MFIs have driven a ordability assessment, robust recov- pand services to a much larger network used close customer contact to achieve single- through remote channels. ery measures and use of partners/channels to digit default rates but this model is expensive maintain customer proximity. One of the key and not scalable. For the virtual micro nance Unlike tradi onal micro-lending, remote chan- bank in context, the client/loan agent ra o steps in our pilot was to ne-tune the auto- nels u lized by a mobile micro nance ins tu- can be almost 5-8 mes that of a tradi onal mated credit risk management ac ons to en- on to deliver credit products can increase the MFI thanks to the higher level of automa- sure low default rates without the need for the MFI/Bank’s por olio at risk. The virtual Micro- on. The lack of close customer contact needs MFI’s direct agents to regularly visit clients. 20 For instance, in Tanzania Vodacom had rela onships with only six dealers who run its en re agent network whereas Safaricom has 300 such rela onships in Kenya.12 Beyond Payments „ Next Genera on Mobile Banking for the Masses
  11. 11. The Solutions: Strategic andOperational Learnings from Pilots In this section we recount the challenges stated above and provide learnings gathered from our pilots along with experiences from other markets. Partnership Set-up#1 Achieving strategic align- pilots when one of the partners has had to put ferences ironed out in parallel, either through ment: Partners should have a valuable me and resources on hold due to pilots or while ge ng the organiza on ready similar appe te for ming and changes in the strategic plan as well as delays for commercial launch. A er the ini al itera- in decision-making from the other party. ons between partners (on market studies andinvestments; they should agree upfront on In addi on, we have iden ed four core stra- strategy design) in our pilot in South East Asia,a high level strategy for four core areas and tegic areas for which partners should achievekick-o execu on/pilot at the earliest. as well as in West Africa, a fully-aligned de- substan al agreement upfront. These include tailed launch strategy was s ll missing. Subse- the scope of ac vi es, target segments, prod-We believe that in order for partners to move uct/services and the distribu on model. The quently, the partners went ahead with a pilot,in tandem, they need rstly to agree on the key ques ons that should be addressed at this with most of the strategy elements detailed ininvestments each one is willing to make in stage are illustrated in Figure 6. This process tandem with opera onal learnings from thedeploying mobile micro nance. In addi on, will allow partners to move in the same direc- eld, thereby facilita ng assessment of the im-upfront agreement is crucial for the launch on as they begin the execu on phase. Each pact of strategic decisions and agreement on ming. We have seen mul ple occasions in our item of strategy can then be detailed and dif- the di erent choices available.Figure 6 *Core areas for high level strategic alignmentSource: Oliver Wyman Strategic Review of Pilots 4 Product & Services Beyond Payments „ Next Genera on Mobile Banking for the Masses 13
  12. 12. #2 Agreement on a pro t mod- In West Africa, by providing commissions suite to allow for one product/activity to el that focuses on both ac- that are proportional to the number and subsidize another. For example, the pricing va ons and usage can be size of transactions, the MFI ensures that scheme in our pilots is designed so that the achieved through on-going upfront dis- agents push for increased usage, doubling commission paid by the MFI to the MNO for activity rates after the new pricing and deposits is partly offset by the withdrawal cussions supported by in-depth custom- agent structure (elaborated in section 6) er valua on exercises; partners might fee charged to the customer. Also, the was adopted. operator may need to adopt a long-term need to subsidize certain ac vi es in re- „ While a threshold-based model23 is view and support certain activities at a turn for others. appropriate in maintaining profitability long-term loss. For instance, in the case for small value transactions, it needs to be of the West Africa pilot, the acquisition of Our experience in Africa as well as in Asia simple enough for users to understand. For validates that the decision on the price point new subscribers is subsidized by the mobile instance, in the case of the pilot in South and split for customer fees and commissions21 network operator, who in turn makes East Asia, the operator adopts the same needs to be decided by on-going communica- profits on subsequent cash-in/cash-out thresholds across all products to ensure on between the two partners un l customer ease of understanding for clients who transactions as illustrated in Figure 7. economics22 are well understood through pilot often subscribe to more than one financial „ Exclusivity in the initial stage is ideal in tests. Business planning and customer valua- product. Percentage-based pricing24 can be helping to protect investments made by on exercises are instrumental in op mizing adopted, provided it is sustainable for low both partners in launching mobile money, pro tability. Some of our ndings include: transaction values. facilitating negotiations on profit sharing, as „ The commission model should be a tool to „ The commissioning model should be well as enabling partners to adopt a long- support both activations as well as usage. formulated as a whole for the entire product term view and commitment. Figure 7 * A balanced profit model that focuses on long term usage enhancement Source: Business planning for pilots in West Africa and South East Asia 21 Service fee charged by the distribu on partner, in this case the MNO, to the lead service provider, in this case the Bank. 22 Transac on volumes, transac on size, wallet size etc. 23 Prices are xed for a range of transac on amounts. 24 When the fee is charged as a % of the transac on amount.14 Beyond Payments „ Next Genera on Mobile Banking for the Masses
  13. 13. Marketing#3 „ 3.1 The product se- quence decision should be driven by the mar-ket’s need gap as well as readiness formobile money e.g. in markets new to It might not be the best thing to launch complex 2nd genera on nancial prod- ucts amid the following condi ons: such as savings and loans. Our pilot in West Africa is one such situa on. While the mo- bile payments product exists in the market, its ac vity rate is very low and hence most low income subscribers are new to it. The MFI involved in the pilot is an exis ng smallthe concept, it is advisable to start with „ Lack of partners with existing 1st generation ins tu on with several tradi onal product1st gen products to allow clients/opera- mobile money products/low mobile money lines currently served through brick andtors to mature before launching 2nd gen activity. mortar branches and legacy systems. It hasservices. „ MFI partners have deep-rooted legacy taken more than 12 months to set-up a mo- systems and processes. bile money distribu on channel for the MFIWhen addressing the sequence of product „ There is a real need for 1st generation services (for loan repayment), given the number ofdeployments, or which service should be owing to high domestic remittance volume complexi es in integra ng new models andlaunched rst from the full bouquet of 1st and relatively few safe and convenient processes into its normal business. An imme-and 2nd genera on mobile money products, channels of money transfer. diate shi to mobile savings or credit wouldseveral market and business condi ons need „ Financial literacy as well as telecom have been quite challenging for the be considered. In Figure 8 we have provid- familiarity of target segments is relativelyed some guidelines that can help achieve the low. On the other hand, 2nd genera on mobileright sequence for product launches. Priori zing 1st genera on products in such money products can be appropriate if: condi ons is preferable to allow the model „ There are mature and medium or large-scale to mature, enable providers to go through mobile money deployments in the market. the learning curve and customers to become „ Partnering MFIs/Banks are relatively accustomed to simple and low-risk services such as bill payments and remi ances be- technologically advanced, and can easily fore being o ered 2nd genera on products, integrate their core banking systems with the MNO’s mobile money platforms. „ The volume of domestic remittances isFigure 8 * relatively low, or there are adequate andConditions for choosing product sequence convenient existing modes of money transfer, thereby reducing the need forSource: Assessment of pilots in West Africa and South East Asia mobile payments. „ Customers are financially literate, as well as comfortable in using mobile value-added services. „ Customers trust transactions through non- branch/banks. „ A favorable regulatory environment that allows testing of innovative delivery models for mobile money. An example of such a market is our pilot in South East Asia. The market has mature mo- bile money services. The partnering MFI is a virtual bank focused only on mobile bank- ing products, with no conven onal banking arm (and hence can integrate easily into the MNO’s exis ng mobile money pla orms). The country already has a regulated and easily ac- cessible network of quasi- nancial pawnshops that support money transfer even in remote areas of the country. Banking penetra on, on the other hand, is low and exis ng MFIs/ru- ral banks are not aggressively pushing savings products. All of these factors come together to make condi ons favorable for launching 2nd genera on products in this market through the selected providers. Beyond Payments „ Next Genera on Mobile Banking for the Masses 15
  14. 14. „ 3.1 (cont’d) developing a suitable Figure 9 * product design: low income segments Heterogeneous occupational sub-segments can be very heterogeneous in their financial needs; selectively targeting in one of the pilot geographies segments and tuning the targeting strategy to income cycles can provide a competitive edge to partners. Segment Size Saving capacity Banking access Speci c needs Savings Loans Manager/Civil Low High High Low Many Many The nancial needs of low income segments servants can be well understood through systema c Professionals Low High High Low Many Many market research and pilot tests. In both our pilots we observed notable di erences in the Clerks/Technicians Low High High Mid Many Many nancial behaviors of various target segments. These di erences relate to the level of nancial Micro entrepreneurs Mid Mid Mid High Some Some exposure, access to banking, savings, and m- Fisherman/ livestock High Mid Low High Few Some ing of high and low income periods. Segments also tend to di er in terms of their occupa onal Farmers High Mid Low High Few Some ecosystems and cash ow needs which in u- ence the nature of the nancial products to be Low wage workers High Low Low Mid Some Few delivered. Partners should iden fy the largest Service employees High Mid Mid Low Many Some sub-segments (by value and volume) and de- sign products based on their typical needs. In Pensioners Low Low Mid Mid Many Few South East Asia, we observed that occupa on greatly a ects the nancial behavior of low-in- Unfavorable for Mobile MFI Slightly favorable for Mobile MFI Very favorable for Mobile MFI come segments. In addi on, we iden ed the need to segment the market geographically, given the country’s regional dispari es as illus- trated in Figure 9 and 10. Partners should also study both magnitude and Figure 10 * cyclical nature of the nancial needs of selected target segments, and design products to ad- Opportunity assessment by regions dress as well as take advantage of such cycles. In one of our pilots we observed that owing to the changes in income pa erns through the Target MNO’s MNO’s network year, it would not be ideal to sell the same - Region Pop. density segment size Bank density coverage coverage nancial product all year long. The level of sav- ings varied signi cantly, becoming nega ve in Region I Very High Low High High Tbd certain months and increasing by 200-300% in Region II Low Low Mid High Tbd others, as illustrated in Figure 11. This called for a exible product design. Such di erences were Region III Mid Mid Mid Mid Tbd not only seen across segments and months but also across di erent regions within the country. Region IV Low High Mid Low Tbd Mee ng these needs did not necessarily require Region V Mid Low High High Tbd a number of di erent products, but only modi- ca ons in the design and marke ng approach Region VI High Low High High Tbd to the core product. We observed that just by reducing the minimum deposit limits, the MFI Region VII Low Mid Low Low Tbd is able to sustain ac vity levels in low income Region VIII Mid High Low High Tbd months, while in high income months it can capture bulk deposits with targeted promo ons. Region IX Mid High Mid High Tbd Even for Safaricom in Kenya, for 1st genera on products like money transfer, the seasonality Region X High Mid Mid High Tbd trends are evident – the ends of the month be- ing higher for cash-in transac ons, and week- Low opportunity Unfavorable for Bank ends higher for cash-out transac ons. In speci c Mid opportunity Slightly favorable for Bank loca ons (market towns), certain agents see a High opportunity Very favorable for Bank high volume of cash-out transac ons in the early morning, while the late a ernoon transac ons are cash-in. Agents learn from these trends to ensure their cash- oat is adequate.16 Beyond Payments „ Next Genera on Mobile Banking for the Masses